Namespace: ttsAdapters

Registry of TTS adapters and related utilities.

Classes

BaseTTSAdapter
GoogleTTSAdapter
RestTTSAdapter
TTSAbortError
TTSSynthesizeError

Methods

(async, static) buildAuth(cfg) → {Promise.<{authHeaders: Object.<string, string>, authQuery: Object.<string, string>}>}

Build Authorization header (via jwtGet) and optional apiKey query.

Parameters:

Name Type Description

cfg

Object

Properties

Name	Type	Attributes	Description
`jwtGet`	function	<optional>	Optional async function to get JWT.
`jwtHeaderName`	string	<optional>	Header name, defaults to "Authorization".
`jwtPrefix`	string	<optional>	Value prefix, defaults to "Bearer ".
`apiKey`	string	<optional>	API key value.
`apiKeyParam`	string	<optional>	API key query parameter name, defaults to "key".

Returns:

Type:: Promise.<{authHeaders: Object.<string, string>, authQuery: Object.<string, string>}>

(static) defaultBuildRequest(input, cfg) → {Object}

Default request builder (HeadTTS-like JSON payload).

Parameters:

Name	Type	Description
`input`	Object
`cfg`	Object

Returns:

Type:: Object

(async, static) defaultParseResponse(res, input, _cfg) → {Promise.<*>}

Default response parser: expects JSON with base64 "audioContent" (or raw audio) and optional timings.

Parameters:

Name	Type	Description
`res`	Response
`input`	Object
`_cfg`	Object

Returns:

Resolves with adapter output (see contracts: TTSAdapterOutput).

Type:: Promise.<*>

(async, static) safeFetch(input, initopt) → {Promise.<Response>}

Wrap fetch with AbortSignal support and uniform errors.

Parameters:

Name	Type	Attributes	Description
`input`	RequestInfo
`init`	RequestInit	<optional>	Fetch options. May include an `AbortSignal` to support cancellation.

Throws:

When aborted via init.signal.
Type TTSAbortError

Returns:

Type:: Promise.<Response>

Type Definitions

TTSAdapter

Minimal TTS adapter interface.

Type:

Object

Properties

Name	Type	Description
`synthesize`	function

TTSAdapterInput

TTS adapter input. Plain text only (no SSML). Words array is optional but recommended if you want 1:1 word timings. The adapter should respect AbortSignal for cancellation.

Type:

Object

Properties

Name	Type	Attributes	Default	Description
`text`	string			plain text to synthesize (no SSML)
`ssml`	string	<optional>		optional SSML; if provided, adapter should prefer it over plain text
`words`	Array.<string>	<optional>		words array (if provided, adapter should return 1:1 timings)
`voice`	string \| Object	<optional>		string or provider-specific voice object
`rate`	number	<optional>		normalized speed, adapter maps to provider scale
`pitch`	number	<optional>		semitones or provider scale
`volume`	number	<optional>		dB or provider scale
`timeUnit`	'ms' \| 's'	<optional>	'ms'	time unit for words timings if provided
`batchMode`	'sentence' \| 'all'	<optional>	'sentence'	how to batch input text
`adapterConfig`	any	<optional>		endpoint, headers, keys, etc.
`signal`	AbortSignal	<optional>		optional signal to support cancellation

TTSAdapterOutput

TTS adapter output. If provider does not support timings — return only audio. If you provide word-level timings use fields: words, wtimes, wduration. If you provide visemes — use fields: visemes, vtimes, vduration.

NOTE: timeUnit defaults to 'ms' unless explicitly set to 's'.

Type:

Object

Properties

Name	Type	Attributes	Default	Description
`audio`	ArrayBuffer	<optional>		raw audio data (e.g., WAV, MP3, OGG)
`timeUnit`	'ms' \| 's'	<optional>	'ms'	time unit for all timings below
`words`	Array.<string>	<optional>		words as returned by provider or echoed from input
`wtimes`	Array.<number>	<optional>		start times of words
`wduration`	Array.<number>	<optional>		durations of words
`visemes`	Array.<string>	<optional>		visemes (e.g., "AI", "E", "U", "O", "CDGKNRSThYZ", "L", "MBP", "FV", "WQ", "rest")
`vtimes`	Array.<number>	<optional>		center times of visemes
`vduration`	Array.<number>	<optional>		durations of visemes